Word embeddings quantify 100 years of gender and ethnic stereotypes

TLDR

Word embeddings encode semantic relationships by representing words as vectors whose geometry reflects meaning. The paper develops a framework to use temporal changes in word embeddings to quantify shifts in stereotypes and attitudes toward women and ethnic minorities in the United States over the 20th and 21st centuries. The framework trains embeddings on a century of text and aligns them with U.S. Census data to track embedding changes alongside demographic and occupational shifts.

Abstract

Word embeddings use vectors to represent words such that the geometry between vectors captures semantic relationship between the words. In this paper, we develop a framework to demonstrate how the temporal dynamics of the embedding can be leveraged to quantify changes in stereotypes and attitudes toward women and ethnic minorities in the 20th and 21st centuries in the United States. We integrate word embeddings trained on 100 years of text data with the U.S. Census to show that changes in the embedding track closely with demographic and occupation shifts over time. The embedding captures global social shifts -- e.g., the women's movement in the 1960s and Asian immigration into the U.S -- and also illuminates how specific adjectives and occupations became more closely associated with certain populations over time. Our framework for temporal analysis of word embedding opens up a powerful new intersection between machine learning and quantitative social science.